Behavior recognition via sparse spatio-temporal features
Research output: Contribution to journal › Conference article › Research › peer-review
Standard
Behavior recognition via sparse spatio-temporal features. / Dollár, Piotr; Rabaud, Vincent; Cottrell, Garrison; Belongie, Serge.
In: Proceedings - 2nd Joint IEEE International Workshop on Visual Surveillance and Performance Evaluation of Tracking and Surveillance, VS-PETS, 2005, p. 65-72.Research output: Contribution to journal › Conference article › Research › peer-review
Harvard
APA
Vancouver
Author
Bibtex
}
RIS
TY - GEN
T1 - Behavior recognition via sparse spatio-temporal features
AU - Dollár, Piotr
AU - Rabaud, Vincent
AU - Cottrell, Garrison
AU - Belongie, Serge
PY - 2005
Y1 - 2005
N2 - A common trend in object recognition is to detect and lever-age the use of sparse, informative feature points, The use of such features makes the problem more manageable while providing increased robustness to noise and pose variation. In this work we develop an extension of these ideas to the spatio-temporal case. For this purpose, we show that the direct 3D counterparts to commonly used 2D interest point detectors are inadequate, and we propose an alternative. Anchoring off of these interest points, we devise a recognition algorithm based on spatio-temporally windowed data. We present recognition results on a variety of datasets including both human and rodent behavior.
AB - A common trend in object recognition is to detect and lever-age the use of sparse, informative feature points, The use of such features makes the problem more manageable while providing increased robustness to noise and pose variation. In this work we develop an extension of these ideas to the spatio-temporal case. For this purpose, we show that the direct 3D counterparts to commonly used 2D interest point detectors are inadequate, and we propose an alternative. Anchoring off of these interest points, we devise a recognition algorithm based on spatio-temporally windowed data. We present recognition results on a variety of datasets including both human and rodent behavior.
UR - http://www.scopus.com/inward/record.url?scp=33846622081&partnerID=8YFLogxK
U2 - 10.1109/VSPETS.2005.1570899
DO - 10.1109/VSPETS.2005.1570899
M3 - Conference article
AN - SCOPUS:33846622081
SP - 65
EP - 72
JO - Proceedings - 2nd Joint IEEE International Workshop on Visual Surveillance and Performance Evaluation of Tracking and Surveillance, VS-PETS
JF - Proceedings - 2nd Joint IEEE International Workshop on Visual Surveillance and Performance Evaluation of Tracking and Surveillance, VS-PETS
T2 - 2nd Joint IEEE International Workshop on Visual Surveillance and Performance Evaluation of Tracking and Surveillance, VS-PETS
Y2 - 15 October 2005 through 16 October 2005
ER -
ID: 302054466